1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=aarch64-none-linux-gnu -mattr=+neon | FileCheck %s
define i1 @combine_setcc_eq_vecreduce_or_v8i1(<8 x i8> %a) {
; CHECK-LABEL: combine_setcc_eq_vecreduce_or_v8i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmeq v0.8b, v0.8b, #0
; CHECK-NEXT: mov w8, #1 // =0x1
; CHECK-NEXT: umaxv b0, v0.8b
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: bic w0, w8, w9
; CHECK-NEXT: ret
%cmp1 = icmp eq <8 x i8> %a, zeroinitializer
%cast = bitcast <8 x i1> %cmp1 to i8
%cmp2 = icmp eq i8 %cast, zeroinitializer
ret i1 %cmp2
}
define i1 @combine_setcc_eq_vecreduce_or_v16i1(<16 x i8> %a) {
; CHECK-LABEL: combine_setcc_eq_vecreduce_or_v16i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmeq v0.16b, v0.16b, #0
; CHECK-NEXT: mov w8, #1 // =0x1
; CHECK-NEXT: umaxv b0, v0.16b
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: bic w0, w8, w9
; CHECK-NEXT: ret
%cmp1 = icmp eq <16 x i8> %a, zeroinitializer
%cast = bitcast <16 x i1> %cmp1 to i16
%cmp2 = icmp eq i16 %cast, zeroinitializer
ret i1 %cmp2
}
define i1 @combine_setcc_eq_vecreduce_or_v32i1(<32 x i8> %a) {
; CHECK-LABEL: combine_setcc_eq_vecreduce_or_v32i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmeq v1.16b, v1.16b, #0
; CHECK-NEXT: mov w8, #1 // =0x1
; CHECK-NEXT: cmeq v0.16b, v0.16b, #0
; CHECK-NEXT: orr v0.16b, v0.16b, v1.16b
; CHECK-NEXT: umaxv b0, v0.16b
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: bic w0, w8, w9
; CHECK-NEXT: ret
%cmp1 = icmp eq <32 x i8> %a, zeroinitializer
%cast = bitcast <32 x i1> %cmp1 to i32
%cmp2 = icmp eq i32 %cast, zeroinitializer
ret i1 %cmp2
}
define i1 @combine_setcc_eq_vecreduce_or_v64i1(<64 x i8> %a) {
; CHECK-LABEL: combine_setcc_eq_vecreduce_or_v64i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmeq v2.16b, v2.16b, #0
; CHECK-NEXT: mov w8, #1 // =0x1
; CHECK-NEXT: cmeq v3.16b, v3.16b, #0
; CHECK-NEXT: cmeq v1.16b, v1.16b, #0
; CHECK-NEXT: cmeq v0.16b, v0.16b, #0
; CHECK-NEXT: orr v1.16b, v1.16b, v3.16b
; CHECK-NEXT: orr v0.16b, v0.16b, v2.16b
; CHECK-NEXT: orr v0.16b, v0.16b, v1.16b
; CHECK-NEXT: umaxv b0, v0.16b
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: bic w0, w8, w9
; CHECK-NEXT: ret
%cmp1 = icmp eq <64 x i8> %a, zeroinitializer
%cast = bitcast <64 x i1> %cmp1 to i64
%cmp2 = icmp eq i64 %cast, zeroinitializer
ret i1 %cmp2
}
define i1 @combine_setcc_ne_vecreduce_or_v8i1(<8 x i8> %a) {
; CHECK-LABEL: combine_setcc_ne_vecreduce_or_v8i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmtst v0.8b, v0.8b, v0.8b
; CHECK-NEXT: umaxv b0, v0.8b
; CHECK-NEXT: fmov w8, s0
; CHECK-NEXT: and w0, w8, #0x1
; CHECK-NEXT: ret
%cmp1 = icmp ne <8 x i8> %a, zeroinitializer
%cast = bitcast <8 x i1> %cmp1 to i8
%cmp2 = icmp ne i8 %cast, zeroinitializer
ret i1 %cmp2
}
define i1 @combine_setcc_ne_vecreduce_or_v16i1(<16 x i8> %a) {
; CHECK-LABEL: combine_setcc_ne_vecreduce_or_v16i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmtst v0.16b, v0.16b, v0.16b
; CHECK-NEXT: umaxv b0, v0.16b
; CHECK-NEXT: fmov w8, s0
; CHECK-NEXT: and w0, w8, #0x1
; CHECK-NEXT: ret
%cmp1 = icmp ne <16 x i8> %a, zeroinitializer
%cast = bitcast <16 x i1> %cmp1 to i16
%cmp2 = icmp ne i16 %cast, zeroinitializer
ret i1 %cmp2
}
define i1 @combine_setcc_ne_vecreduce_or_v32i1(<32 x i8> %a) {
; CHECK-LABEL: combine_setcc_ne_vecreduce_or_v32i1:
; CHECK: // %bb.0:
; CHECK-NEXT: orr v0.16b, v0.16b, v1.16b
; CHECK-NEXT: cmtst v0.16b, v0.16b, v0.16b
; CHECK-NEXT: umaxv b0, v0.16b
; CHECK-NEXT: fmov w8, s0
; CHECK-NEXT: and w0, w8, #0x1
; CHECK-NEXT: ret
%cmp1 = icmp ne <32 x i8> %a, zeroinitializer
%cast = bitcast <32 x i1> %cmp1 to i32
%cmp2 = icmp ne i32 %cast, zeroinitializer
ret i1 %cmp2
}
define i1 @combine_setcc_ne_vecreduce_or_v64i1(<64 x i8> %a) {
; CHECK-LABEL: combine_setcc_ne_vecreduce_or_v64i1:
; CHECK: // %bb.0:
; CHECK-NEXT: orr v1.16b, v1.16b, v3.16b
; CHECK-NEXT: orr v0.16b, v0.16b, v2.16b
; CHECK-NEXT: orr v0.16b, v0.16b, v1.16b
; CHECK-NEXT: cmtst v0.16b, v0.16b, v0.16b
; CHECK-NEXT: umaxv b0, v0.16b
; CHECK-NEXT: fmov w8, s0
; CHECK-NEXT: and w0, w8, #0x1
; CHECK-NEXT: ret
%cmp1 = icmp ne <64 x i8> %a, zeroinitializer
%cast = bitcast <64 x i1> %cmp1 to i64
%cmp2 = icmp ne i64 %cast, zeroinitializer
ret i1 %cmp2
}
define i1 @combine_setcc_eq_vecreduce_and_v8i1(<8 x i8> %a) {
; CHECK-LABEL: combine_setcc_eq_vecreduce_and_v8i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmeq v0.8b, v0.8b, #0
; CHECK-NEXT: uminv b0, v0.8b
; CHECK-NEXT: fmov w8, s0
; CHECK-NEXT: and w0, w8, #0x1
; CHECK-NEXT: ret
%cmp1 = icmp eq <8 x i8> %a, zeroinitializer
%cast = bitcast <8 x i1> %cmp1 to i8
%cmp2 = icmp eq i8 %cast, -1
ret i1 %cmp2
}
define i1 @combine_setcc_eq_vecreduce_and_v16i1(<16 x i8> %a) {
; CHECK-LABEL: combine_setcc_eq_vecreduce_and_v16i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmeq v0.16b, v0.16b, #0
; CHECK-NEXT: uminv b0, v0.16b
; CHECK-NEXT: fmov w8, s0
; CHECK-NEXT: and w0, w8, #0x1
; CHECK-NEXT: ret
%cmp1 = icmp eq <16 x i8> %a, zeroinitializer
%cast = bitcast <16 x i1> %cmp1 to i16
%cmp2 = icmp eq i16 %cast, -1
ret i1 %cmp2
}
define i1 @combine_setcc_eq_vecreduce_and_v32i1(<32 x i8> %a) {
; CHECK-LABEL: combine_setcc_eq_vecreduce_and_v32i1:
; CHECK: // %bb.0:
; CHECK-NEXT: orr v0.16b, v0.16b, v1.16b
; CHECK-NEXT: cmeq v0.16b, v0.16b, #0
; CHECK-NEXT: uminv b0, v0.16b
; CHECK-NEXT: fmov w8, s0
; CHECK-NEXT: and w0, w8, #0x1
; CHECK-NEXT: ret
%cmp1 = icmp eq <32 x i8> %a, zeroinitializer
%cast = bitcast <32 x i1> %cmp1 to i32
%cmp2 = icmp eq i32 %cast, -1
ret i1 %cmp2
}
define i1 @combine_setcc_eq_vecreduce_and_v64i1(<64 x i8> %a) {
; CHECK-LABEL: combine_setcc_eq_vecreduce_and_v64i1:
; CHECK: // %bb.0:
; CHECK-NEXT: orr v1.16b, v1.16b, v3.16b
; CHECK-NEXT: orr v0.16b, v0.16b, v2.16b
; CHECK-NEXT: orr v0.16b, v0.16b, v1.16b
; CHECK-NEXT: cmeq v0.16b, v0.16b, #0
; CHECK-NEXT: uminv b0, v0.16b
; CHECK-NEXT: fmov w8, s0
; CHECK-NEXT: and w0, w8, #0x1
; CHECK-NEXT: ret
%cmp1 = icmp eq <64 x i8> %a, zeroinitializer
%cast = bitcast <64 x i1> %cmp1 to i64
%cmp2 = icmp eq i64 %cast, -1
ret i1 %cmp2
}
define i1 @combine_setcc_ne_vecreduce_and_v8i1(<8 x i8> %a) {
; CHECK-LABEL: combine_setcc_ne_vecreduce_and_v8i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmtst v0.8b, v0.8b, v0.8b
; CHECK-NEXT: mov w8, #1 // =0x1
; CHECK-NEXT: uminv b0, v0.8b
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: bic w0, w8, w9
; CHECK-NEXT: ret
%cmp1 = icmp ne <8 x i8> %a, zeroinitializer
%cast = bitcast <8 x i1> %cmp1 to i8
%cmp2 = icmp ne i8 %cast, -1
ret i1 %cmp2
}
define i1 @combine_setcc_ne_vecreduce_and_v16i1(<16 x i8> %a) {
; CHECK-LABEL: combine_setcc_ne_vecreduce_and_v16i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmtst v0.16b, v0.16b, v0.16b
; CHECK-NEXT: mov w8, #1 // =0x1
; CHECK-NEXT: uminv b0, v0.16b
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: bic w0, w8, w9
; CHECK-NEXT: ret
%cmp1 = icmp ne <16 x i8> %a, zeroinitializer
%cast = bitcast <16 x i1> %cmp1 to i16
%cmp2 = icmp ne i16 %cast, -1
ret i1 %cmp2
}
define i1 @combine_setcc_ne_vecreduce_and_v32i1(<32 x i8> %a) {
; CHECK-LABEL: combine_setcc_ne_vecreduce_and_v32i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmtst v0.16b, v0.16b, v0.16b
; CHECK-NEXT: mov w8, #1 // =0x1
; CHECK-NEXT: cmeq v1.16b, v1.16b, #0
; CHECK-NEXT: bic v0.16b, v0.16b, v1.16b
; CHECK-NEXT: uminv b0, v0.16b
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: bic w0, w8, w9
; CHECK-NEXT: ret
%cmp1 = icmp ne <32 x i8> %a, zeroinitializer
%cast = bitcast <32 x i1> %cmp1 to i32
%cmp2 = icmp ne i32 %cast, -1
ret i1 %cmp2
}
define i1 @combine_setcc_ne_vecreduce_and_v64i1(<64 x i8> %a) {
; CHECK-LABEL: combine_setcc_ne_vecreduce_and_v64i1:
; CHECK: // %bb.0:
; CHECK-NEXT: cmtst v1.16b, v1.16b, v1.16b
; CHECK-NEXT: mov w8, #1 // =0x1
; CHECK-NEXT: cmtst v0.16b, v0.16b, v0.16b
; CHECK-NEXT: cmeq v3.16b, v3.16b, #0
; CHECK-NEXT: cmeq v2.16b, v2.16b, #0
; CHECK-NEXT: bic v1.16b, v1.16b, v3.16b
; CHECK-NEXT: bic v0.16b, v0.16b, v2.16b
; CHECK-NEXT: and v0.16b, v0.16b, v1.16b
; CHECK-NEXT: uminv b0, v0.16b
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: bic w0, w8, w9
; CHECK-NEXT: ret
%cmp1 = icmp ne <64 x i8> %a, zeroinitializer
%cast = bitcast <64 x i1> %cmp1 to i64
%cmp2 = icmp ne i64 %cast, -1
ret i1 %cmp2
}
define i1 @combine_setcc_eq0_conjunction_xor_or(ptr %a, ptr %b) {
; CHECK-LABEL: combine_setcc_eq0_conjunction_xor_or:
; CHECK: // %bb.0:
; CHECK-NEXT: ldp x8, x9, [x0]
; CHECK-NEXT: ldp x10, x11, [x1]
; CHECK-NEXT: cmp x8, x10
; CHECK-NEXT: ccmp x9, x11, #0, eq
; CHECK-NEXT: cset w0, eq
; CHECK-NEXT: ret
%bcmp = tail call i32 @bcmp(ptr dereferenceable(16) %a, ptr dereferenceable(16) %b, i64 16)
%cmp = icmp eq i32 %bcmp, 0
ret i1 %cmp
}
define i1 @combine_setcc_ne0_conjunction_xor_or(ptr %a, ptr %b) {
; CHECK-LABEL: combine_setcc_ne0_conjunction_xor_or:
; CHECK: // %bb.0:
; CHECK-NEXT: ldp x8, x9, [x0]
; CHECK-NEXT: ldp x10, x11, [x1]
; CHECK-NEXT: cmp x8, x10
; CHECK-NEXT: ccmp x9, x11, #0, eq
; CHECK-NEXT: cset w0, ne
; CHECK-NEXT: ret
%bcmp = tail call i32 @bcmp(ptr dereferenceable(16) %a, ptr dereferenceable(16) %b, i64 16)
%cmp = icmp ne i32 %bcmp, 0
ret i1 %cmp
}
; Doesn't increase the number of instructions, where the LHS has multiple uses
define i32 @combine_setcc_multiuse(i32 %0, i32 %1, i32 %2, i32 %3) {
; CHECK-LABEL: combine_setcc_multiuse:
; CHECK: // %bb.0:
; CHECK-NEXT: eor w8, w1, w0
; CHECK-NEXT: eor w9, w3, w2
; CHECK-NEXT: orr w8, w9, w8
; CHECK-NEXT: cbz w8, .LBB18_2
; CHECK-NEXT: // %bb.1:
; CHECK-NEXT: mov w0, w8
; CHECK-NEXT: b use
; CHECK-NEXT: .LBB18_2:
; CHECK-NEXT: ret
%5 = xor i32 %1, %0
%6 = xor i32 %3, %2
%7 = or i32 %6, %5
%8 = icmp eq i32 %7, 0
br i1 %8, label %11, label %9
9: ; preds = %4
%10 = tail call i32 @use(i32 %7) #2
br label %11
11: ; preds = %4, %9
%12 = phi i32 [ %10, %9 ], [ %0, %4 ]
ret i32 %12
}
; There may be issues with the CMP/CCMP with the scheduling of instructions
; that ISel will create out of the DAG
define i32 @combine_setcc_glue(i128 noundef %x, i128 noundef %y) {
; CHECK-LABEL: combine_setcc_glue:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: cmp x0, x2
; CHECK-NEXT: cset w0, eq
; CHECK-NEXT: ret
entry:
%cmp3 = icmp eq i128 %x, %y
%conv = trunc i128 %x to i64
%conv1 = trunc i128 %y to i64
%cmp = icmp eq i64 %conv, %conv1
%or7 = or i1 %cmp3, %cmp
%or = zext i1 %or7 to i32
ret i32 %or
}
; Reduced test from https://github.com/llvm/llvm-project/issues/58675
define [2 x i64] @PR58675(i128 %a.addr, i128 %b.addr) {
; CHECK-LABEL: PR58675:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: mov x8, xzr
; CHECK-NEXT: mov x9, xzr
; CHECK-NEXT: .LBB20_1: // %do.body
; CHECK-NEXT: // =>This Inner Loop Header: Depth=1
; CHECK-NEXT: cmp x0, x8
; CHECK-NEXT: sbcs xzr, x1, x9
; CHECK-NEXT: csel x10, x0, x8, lo
; CHECK-NEXT: csel x11, x1, x9, lo
; CHECK-NEXT: subs x8, x2, x10
; CHECK-NEXT: sbc x9, x3, x11
; CHECK-NEXT: cmp x3, x11
; CHECK-NEXT: ccmp x2, x10, #0, eq
; CHECK-NEXT: b.ne .LBB20_1
; CHECK-NEXT: // %bb.2: // %do.end
; CHECK-NEXT: mov x0, xzr
; CHECK-NEXT: mov x1, xzr
; CHECK-NEXT: ret
entry:
br label %do.body
do.body: ; preds = %do.body, %entry
%a.addr.i1 = phi i128 [ 1, %do.body ], [ 0, %entry ]
%b.addr.i2 = phi i128 [ %sub, %do.body ], [ 0, %entry ]
%0 = tail call i128 @llvm.umin.i128(i128 %a.addr, i128 %b.addr.i2)
%1 = tail call i128 @llvm.umax.i128(i128 0, i128 %a.addr)
%sub = sub i128 %b.addr, %0
%cmp18.not = icmp eq i128 %b.addr, %0
br i1 %cmp18.not, label %do.end, label %do.body
do.end: ; preds = %do.body
ret [2 x i64] zeroinitializer
}
declare i128 @llvm.umin.i128(i128, i128)
declare i128 @llvm.umax.i128(i128, i128)
declare i32 @bcmp(ptr nocapture, ptr nocapture, i64)
declare i32 @use(i32 noundef)
|
